Design And Implementation Of SAR Imaging Algorithm Based On GPU

Synthetic Aperture Radar(SAR) is a new radar technology that is developed in recent decades. It can obtain high-resolution radar images without the influence of illumination and weather.It is widely used in military, industry, agriculture, marine exploration, geological detection and other fields.In order to get images from swath, we need to deal with a lot of echo data. The essence of SAR imaging is using signal-processing method to handle these data and then obtain high-resolution images. SAR imaging algorithms is complex, it requires plenty number of operations. Traditional SAR imaging algorithm based on CPU is weak in real-time, thus, people start to search for new signal processing technology to meet real-time requirements.It has been a decade after NVIDIA launched the first GPU. During this period, GPU is no longer confined to the scope of image processing and display, especially when NVIDIA provided the concept of CUDA(Compute Unified Device Architecture), a new architecture of GPU. People initially realize the power of GPU to make high-speed parallel computing, and, at the same time, lead a worldwide scientific research trend on using GPU to accelerate scientific computing algorithm. SAR workers are also aware of the advantages of this technology, they are trying to put SAR imaging algorithms on GPU, so that they could meet their needs, shorten imaging time.This paper focus on the applications of SAR imaging on GPU, designed SAR imaging algorithms on GPU, which make use of GPU parallel computing advantages, shorten SAR imaging time and enhance performance in simulation system.This paper covers the following contents: Introduce the SAR imaging principle and how to obtain high-resolution radar images via pulse compression technology. It gives imaging algorithms that deal with positive side model, strabismus and oblique model, and compares these three models. Describe the technical features of GPU and design SAR imaging algorithms and SAR simulation system based on it, the system input-parameters of the radar presents in master interface and starts imaging algorithms and shows the final results on the interface.The main contribution and innovation of this paper are:1. Designed the SAR imaging algorithm which is applied to the model strabismus SAR. 2. Designed the SAR imaging algorithm which is applied to the model oblique SAR. 3. Designed these three SAR imaging algorithms based on GPU. 4. Designed the SAR imaging simulation system based on GPU.